#snark Community Insights & Market Sentiment

📰 What Vitalik discussed this time was not short-term market trends, but whether the computational cost of cryptography can keep being pushed down.

He believes that SNARKs, fully homomorphic encryption (FHE), and indistinguishable obfuscation (iO) all have a chance to bring the extra overhead below 10x by the end of this decade. This is calculated not from the price of a single piece of hardware, but from energy costs plus amortized computation costs.

🔥 The more aggressive judgment is that each of the three has a 33% chance of reaching the "1+ε" level. In other words, when the scale of computation is large enough, the cost added by encrypted processing may gradually approach the cost of the original computation itself.

💡 For all three technologies to get below 10x overhead, he gives it a 60% probability. But the one most likely to first achieve single-digit overhead, Vitalik thinks, may be SNARKs.

To be honest, this judgment is still his own non-consensus view, not something that has already been realized. Dedicated hash functions and some large language model inference have already reached a similar level, which at least shows that this path is not entirely confined to paper.

🤔 If SNARKs really push computational overhead down to single-digit multiples by the end of this decade, what do you think will change first: blockchain verification, or privacy-preserving computing?

#SNARK #以太坊 #密码学 #隐私计算

snark

Trending Topics